1
$\begingroup$

Essentially, I am asking if there have been any attempts at actually simulating how economic agents act in an economic environment and how prices are determined.

To get a better feeling of what I mean, there is this video on the YouTube channel Primer where the YouTuber goes step-by-step explaining how supply and demand can be simulated. He briefly goes over the logic behind how an actor would work, saying that there is a price expectation and a price minimum/maximum (depending if you are a buyer/seller). I couldn't find any information on the code he used for the simulation, but I tried looking for other simulations.

The most promising thing I found was this paper titled "Emergent Economies for Role Playing Games" that tries to simulate a market of different goods with a variety of different buyers and sellers with their own professions with different needs. I found two rudimentary implementations of this paper, bazaarBot by larsiusprime and EconSim by omikun. Both of these implementations wanted to expand further than the simple five commodity economy with a handful of professions by adding in the affect of taxes, banks, trade, investment, currencies, and the likes, but it seems that none of them really updated beyond that.

After this paper I decided I wanted to see if mainstream academia had any more detailed simulations available. At the very least I wanted a model. Unfortunately, I couldn't really find anything. I found a whole bunch of trade models like the Dixit-Stiglitz-Krugman trade model, but I neither thought it was relevant (thought it really only applies to international trade) and I couldn't understand it.

The way I see it is that if you get the basic logic down, a lot of the more complicated models are not needed as they will emerge from the basic actions of individuals. The issue is that (a) I feel a lot of exchange is very subjective (like, for a consumer good, how does the consumer decide what their utility is for that good?) and there is a lot of assumptions that might not be realistic to assume (like information parity among economic agents, and easy entry/exit into markets).

Of course, there is the labour theory of value, but I don't think many mainstream economists give that much thought.

So, I would like to make clear what I am asking:

  1. Are there any research papers/resources that discuss how to simulate the logic an economic agent would need in order for true market behaviour to emerge?

  2. Have there been any implementations? If so, how accurate are they at simulating market behaviour?

$\endgroup$
5
  • 1
    $\begingroup$ You use the word "simulate", but just like in physics, one cannot accurately do that in economics. One can "model", and leave away aspects. Clearly, you have to specify what your goal is, which aspects are important and should not be left out. Even then, the model will not be entirely accurate. So please, let us know what your goa is, what it is that are trying to do with the "simulation". If you want a near-perfect simulation of the general economic decision-making of a human mind, know that neither economics nor neurology is there yet. $\endgroup$
    – Giskard
    Jul 3 at 6:32
  • $\begingroup$ I believe I made it perfectly clear exactly what I want. I don't seek to simulate the world perfectly; I want to see if there have been any actual academic discussions on how to simulate exchange, and how successful models of them are at reflecting how the real world works. It really is no different than me asking if there have been any models on international trade and how successful they are at describing the world. No model of international trade replicates EXACTLY how trade works, but rather to tries to approximate it as accurately as feasible. I am asking if there are any such models. $\endgroup$ Jul 3 at 17:05
  • $\begingroup$ AFAIK just like "general physics models" are not a thing, "general exchange models" are not a thing. Exchanges in financial markets are very different from business-to-business transactions, and business-to-consumer transactions are yet again different. There is also huge variation according to the type of goods. Without narrowing your scope it is likely that you will not get a proper answer. (Declaring your goals and explaining what exactly you mean by the word "simulate", which strikes me as being different from model, may also help.) $\endgroup$
    – Giskard
    Jul 4 at 12:03
  • $\begingroup$ Let's say I wanted to create a game where I am the government of a country and I have access to many government policies that affect the market economy of the country like taxes or price controls. In this game, there would be economic agents that work in a profession to supply goods/services and they themselves in return consume goods/services. To create the game, I would need to determine the internal logic of these agents such that they replicate to an "acceptable" degree how a real market would function which includes responding to price signals in a variety of methods. $\endgroup$ Jul 4 at 19:46
  • $\begingroup$ I disagree that realism is necessary here. In this case your time is probably better spent reading about the mechanisms in the games of the Paradox Interactive game studio. These are highly complex and much enjoyed by all, despite having some significant bugs and having never been tested for accuracy. $\endgroup$
    – Giskard
    Jul 4 at 20:47

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.